Conversation

Laura Deming
@LauraDeming
I think one of my major frustrations with most scientific papers is they don’t take the semantic concept graph seriously. They treat the summary and conclusion as a place to write marketing material, not to seriously describe the knowledge work which motivated the experiment.
14
16
268
Laura Deming
@LauraDeming
This is not taking conceptual work seriously - one of the failures of modern biology is we don’t have a strong theoretical practice, and don’t understand how to value this work. "It should be obvious what data to collect, and how to interpret the results". This is utter baloney.
2
4
41
Laura Deming
@LauraDeming
Not externalizing this work means a massive loss in creative potential, as everyone attempts to recreate the same semantic graphs from scratch, in parallel.
4
2
44
Andy Matuschak
@andy_matuschak
This seems to be a huge problem In HCI, too. “Vision papers” are more or less taboo, so bread-and-butter papers are all grasping at conceptual aims which remain implicit and poorly-understood (including by the investigators, I feel).
1
7
Andy Matuschak
@andy_matuschak
My sense is that HCI wants to think of itself as an empirical science and is sort of nervous about being so pre-Newtonian; vision papers acutely speak to a non-positivist epistemology.
1
2
José Luis Ricón Fernández de la Puente
@ArtirKel
Do you see in there situations of Person A wants to make it concrete and Person B pushing back saying they are misinterpreting their vision and that it won't work, but not giving another concrete impl., saying that it still needs more research?
1
Andy Matuschak
@andy_matuschak
Possibly, yes! This type of conceptual contribution is a component of “vision.” Another missing component is: what problems should we solve, and why? My sense is that TCS is better calibrated there.